In this article, I generate time series data of sales and visualize sales by month and day of the week.
※If you have a data, you don't have to read contents1,2
Contents
1. Creation of simuration data 1
2. Creation of simuration data 2
3. Data combination
4. Organize time series data
5. Total sales by day of the week and month
6. Data visualization
1. Creation of simulation data 1
> time <- seq(as.Date("2020/03/01"), as.Date("2020/03/30"), "day")
> sales <- rpois(length(time), lambda=10)
> Data1 <- data.frame(matrix(NA, ncol=2,nrow=length(time)))
> names(Data1) <- c("time", "sales")
> Data1$time <- time
> Data1$sales <- sales
> head(Data1)
time sales
1 2020-03-01 10
2 2020-03-02 10
3 2020-03-03 10
4 2020-03-04 6
5 2020-03-05 14
6 2020-03-06 12
2. Creation of simulation data 2
> time <- seq(as.Date("2020/04/01"), as.Date("2020/04/30"), "day")
> sales <- rpois(length(time), lambda=10)
> Data2 <- data.frame(matrix(NA, ncol=2,nrow=length(time)))
> names(Data2) <- c("time", "sales")
> Data2$time <- time
> Data2$sales <- sales
> head(Data2)
time sales
1 2020-04-01 15
2 2020-04-02 12
3 2020-04-03 13
4 2020-04-04 7
5 2020-04-05 11
6 2020-04-06 11
3. Data combination
> nrow(Data1)
[1] 30
> nrow(Data2)
[1] 30
> Data <- rbind(Data1, Data2)
> nrow(Data)
[1] 60
4. Organize time series data
> library(lubridate)
> Data$year <- year(Data$time)
> Data$month <- month(Data$time)
> Data$week1 <- wday(Data$time)
> Data$week2 <- weekdays(Data$time)
> Data$time <- as.POSIXlt(Data$time)
> Data$date <- Data$time$mday
> head(Data,8)
time sales year month week1 week2 date
1 2020-03-01 10 2020 3 1 日曜日 1
2 2020-03-02 10 2020 3 2 月曜日 2
3 2020-03-03 10 2020 3 3 火曜日 3
4 2020-03-04 6 2020 3 4 水曜日 4
5 2020-03-05 14 2020 3 5 木曜日 5
6 2020-03-06 12 2020 3 6 金曜日 6
7 2020-03-07 11 2020 3 7 土曜日 7
8 2020-03-08 9 2020 3 1 日曜日 8
5. Total sales by day of the week and month
> weekly_sales <- aggregate(Data$sales, list(Data$week1), sum)
> weekly_sales$Group.1 <- c("日", "月", "火", "水", "木", "金", "土")
> weekly_sales
Group.1 x
1 日 90
2 月 118
3 火 76
4 水 95
5 木 83
6 金 96
7 土 76
> monthly_sales <- aggregate(Data$sales, list(Data$month), sum)
> monthly_sales$Group.1 <- c("3月", "4月")
> monthly_sales
Group.1 x
1 3月 331
2 4月 303
6. Data visualization
> library(ggplot2)
> p1 <- ggplot(data=weekly_sales, aes(x=Group.1))+
+ geom_bar(aes(y=x), size=1.8, stat="identity" )+
+ labs(title="Sales")+
+ ylab("Weekly sales")
>
> p2 <- ggplot(data=monthly_sales, aes(x=Group.1))+
+ geom_bar(aes(y=x), size=1.8, stat="identity" )+
+ labs(title="Sales")+
+ ylab("month sales")
>
> gridExtra::grid.arrange(p1, p2)
